import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

Data Collection

This dataset was downloaded from Kaggle.
(Link to Kaggle dataset): https://www.kaggle.com/code/marto24/bankruptcy-detection

data = pd.read_csv('data.csv')
data = data.dropna()

Exploratory Data Analysis

Looking at the class frequencies to look at the distribution of the data

data['Bankrupt?'].value_counts(normalize= True).plot(kind= 'bar')
plt.xlabel("Bankrupt classes")
plt.ylabel("Frequancy")
plt.title("Class balance")

Text(0.5, 1.0, 'Class balance')

The data seems to be very imbalanced. Checking the specific distributive percentages of the classes.

print(data['Bankrupt?'].value_counts(normalize= True))

Bankrupt?
0    0.967737
1    0.032263
Name: proportion, dtype: float64

Correlation Matrix

plt.figure(figsize=(200, 160))  # Set the figure size in inches (width, height)
corr = data.corr()
sns.heatmap(corr, annot = True, vmin=-1, vmax=1, cmap = 'RdBu', linewidths=0.5)

<Axes: >

The high number of variables make the correlation matrix hard to analyze, but it is quite clear that there are some variables that contribute more significantly to bankruptcy and are scarcely distributed. Most of the dataset consists of variables that have relatively trivial contributions to Bankruptcy.
To look at the variables that have non-trivial contributions, we set a threshold of $0.2$.

vars = []

for i in range(len(corr["Bankrupt?"])):
    if abs(corr["Bankrupt?"][i]) > 0.15:
        vars.append(corr["Bankrupt?"].index[i])

print(vars)

['Bankrupt?', ' ROA(C) before interest and depreciation before interest', ' ROA(A) before interest and % after tax', ' ROA(B) before interest and depreciation after tax', ' Net Value Per Share (B)', ' Net Value Per Share (A)', ' Net Value Per Share (C)', ' Persistent EPS in the Last Four Seasons', ' Per Share Net profit before tax (Yuan ¥)', ' Debt ratio %', ' Net worth/Assets', ' Borrowing dependency', ' Net profit before tax/Paid-in capital', ' Working Capital to Total Assets', ' Current Liability to Assets', ' Current Liabilities/Equity', ' Retained Earnings to Total Assets', ' Current Liability to Equity', ' Current Liability to Current Assets', ' Net Income to Total Assets', " Net Income to Stockholder's Equity", ' Liability to Equity']

/var/folders/f9/jdxhns6s5xqbjc5h71x7s2740000gn/T/ipykernel_3638/1875777422.py:4: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`
  if abs(corr["Bankrupt?"][i]) > 0.15:

Looking at the correlation matrix, we have stored the variables that have a non-trivial correlation $( > 0.2)$ to bankruptcy in the variable var.

Lets look at the distributions of the filtered variables.

fig, axes = plt.subplots(len(vars) // 3, 3, figsize=(20, 40))
i, j = 0, 0
for var in vars:
    if var == "Bankrupt?":
        continue
    sns.boxenplot(ax = axes[i, j], x="Bankrupt?", y=var, data=data, palette="Blues", hue = "Bankrupt?")
    plt.xlabel("Bankrupt classes")
    plt.ylabel(var, fontsize= 1)
    j += 1
    if j == 3:
        j = 0
        i += 1

While the plots could be informative, the significant presence of outliers makes anaysis quite hard. This warrants filtering the outliers and scaling the data to ensure a more comprehensive and representative plot.

Data Cleaning and Preprocessing

Filtering the data by the variables that have a non-trivial correlation to Bankruptcy

data = data[vars]
display(data.head())

	Bankrupt?	ROA(C) before interest and depreciation before interest	ROA(A) before interest and % after tax	ROA(B) before interest and depreciation after tax	Net Value Per Share (B)	Net Value Per Share (A)	Net Value Per Share (C)	Persistent EPS in the Last Four Seasons	Per Share Net profit before tax (Yuan ¥)	Debt ratio %	...	Net profit before tax/Paid-in capital	Working Capital to Total Assets	Current Liability to Assets	Current Liabilities/Equity	Retained Earnings to Total Assets	Current Liability to Equity	Current Liability to Current Assets	Net Income to Total Assets	Net Income to Stockholder's Equity	Liability to Equity
0	1	0.370594	0.424389	0.405750	0.147950	0.147950	0.147950	0.169141	0.138736	0.207576	...	0.137757	0.672775	0.147308	0.339077	0.903225	0.339077	0.118250	0.716845	0.827890	0.290202
1	1	0.464291	0.538214	0.516730	0.182251	0.182251	0.182251	0.208944	0.169918	0.171176	...	0.168962	0.751111	0.056963	0.329740	0.931065	0.329740	0.047775	0.795297	0.839969	0.283846
2	1	0.426071	0.499019	0.472295	0.177911	0.177911	0.193713	0.180581	0.142803	0.207516	...	0.148036	0.829502	0.098162	0.334777	0.909903	0.334777	0.025346	0.774670	0.836774	0.290189
3	1	0.399844	0.451265	0.457733	0.154187	0.154187	0.154187	0.193722	0.148603	0.151465	...	0.147561	0.725754	0.098715	0.331509	0.906902	0.331509	0.067250	0.739555	0.834697	0.281721
4	1	0.465022	0.538432	0.522298	0.167502	0.167502	0.167502	0.212537	0.168412	0.106509	...	0.167461	0.751822	0.110195	0.330726	0.913850	0.330726	0.047725	0.795016	0.839973	0.278514

5 rows × 22 columns

Removing outliers using the IQR method
Scaling the data using the Standard Scaler

from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split

# Check for missing n if theree is fill with median
missing_values = data.isnull().sum()
data.fillna(data.median(), inplace=True)
data = data.drop_duplicates()


#removes outliers


for column in data.columns[1:]:  # Skip the first column as it's the target variable
        df = data
        Q1 = df[column].quantile(0.25)
        Q3 = df[column].quantile(0.75)
        IQR = Q3 - Q1
        lower_limit = Q1 - 1.5 * IQR
        upper_limit = Q3 + 1.5 * IQR
        df[column] = df[column].clip(lower_limit, upper_limit)


#feature scaling

scaler = StandardScaler()
scaled_features = scaler.fit_transform(data.drop(columns=['Bankrupt?']))
scaled_df = pd.DataFrame(scaled_features, columns=data.columns[1:])

We can now go back to our boxplots to look at the distributions of the variables that have a non-trivial correlation to bankruptcy.

fig, axes = plt.subplots(len(vars) // 3, 3, figsize=(20, 40))
i, j = 0, 0
for var in vars:
    if var == "Bankrupt?":
        continue
    sns.boxenplot(ax = axes[i, j], x="Bankrupt?", y=var, data=data, palette="Blues", hue = "Bankrupt?")
    plt.xlabel("Bankrupt classes")
    plt.ylabel(var, fontsize= 1)
    j += 1
    if j == 3:
        j = 0
        i += 1

Now with a more representative visualization of the data distribution, we can see that the distributions of the variables in the negative case of the Bankruptcy variable ("Bankruptcy?" = 0) are normally distrbuted while that of the positive case ("Bankruptcy?" = 1) are skewed.

Data Splitting and Model Selection

We can now split the data into testing and training sets

X = scaled_df
y = data['Bankrupt?']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=21)


df.head()

	Bankrupt?	ROA(C) before interest and depreciation before interest	ROA(A) before interest and % after tax	ROA(B) before interest and depreciation after tax	Net Value Per Share (B)	Net Value Per Share (A)	Net Value Per Share (C)	Persistent EPS in the Last Four Seasons	Per Share Net profit before tax (Yuan ¥)	Debt ratio %	...	Net profit before tax/Paid-in capital	Working Capital to Total Assets	Current Liability to Assets	Current Liabilities/Equity	Retained Earnings to Total Assets	Current Liability to Equity	Current Liability to Current Assets	Net Income to Total Assets	Net Income to Stockholder's Equity	Liability to Equity
0	1	0.387973	0.455122	0.442034	0.147950	0.147950	0.147950	0.178548	0.138736	0.207576	...	0.137757	0.672775	0.147308	0.338662	0.910524	0.338662	0.068886	0.752193	0.836752	0.288207
1	1	0.464291	0.538214	0.516730	0.182251	0.182251	0.182251	0.208944	0.169918	0.171176	...	0.168962	0.751111	0.056963	0.329740	0.931065	0.329740	0.047775	0.795297	0.839969	0.283846
2	1	0.426071	0.499019	0.472295	0.177911	0.177911	0.193713	0.180581	0.142803	0.207516	...	0.148036	0.829502	0.098162	0.334777	0.910524	0.334777	0.025346	0.774670	0.836774	0.288207
3	1	0.399844	0.455122	0.457733	0.154187	0.154187	0.154187	0.193722	0.148603	0.151465	...	0.147561	0.725754	0.098715	0.331509	0.910524	0.331509	0.067250	0.752193	0.836752	0.281721
4	1	0.465022	0.538432	0.522298	0.167502	0.167502	0.167502	0.212537	0.168412	0.106509	...	0.167461	0.751822	0.110195	0.330726	0.913850	0.330726	0.047725	0.795016	0.839973	0.278514

5 rows × 22 columns

Looking at various metrics to determine the best model for the data

ROC AUC

from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn.ensemble import AdaBoostClassifier
from sklearn.model_selection import cross_val_score
from sklearn.neural_network import MLPClassifier
from sklearn.metrics import accuracy_score, roc_auc_score, classification_report

models = {
    "Logistic Regression": LogisticRegression(max_iter=1000),
    "Decision Tree": DecisionTreeClassifier(criterion='entropy'),
    "Random Forest": RandomForestClassifier(),
    "Gradient Boosting": GradientBoostingClassifier(),
    "Support Vector Machine": SVC(probability=True),
    "Neural Network": MLPClassifier(learning_rate='adaptive', max_iter=1000),
    "XGBoost": XGBClassifier(),
    "AdaBoost": AdaBoostClassifier()
}

results = {}
for name, model in models.items():
    cv_scores = cross_val_score(model, X_train, y_train, cv=5, scoring='roc_auc')
    results[name] = cv_scores.mean()

for model, score in results.items():
    print(f"{model}: Mean ROC AUC = {score:.4f}")

Logistic Regression: Mean ROC AUC = 0.9325
Decision Tree: Mean ROC AUC = 0.6276
Random Forest: Mean ROC AUC = 0.9055
Gradient Boosting: Mean ROC AUC = 0.9208
Support Vector Machine: Mean ROC AUC = 0.8028
Neural Network: Mean ROC AUC = 0.9182
XGBoost: Mean ROC AUC = 0.9113
AdaBoost: Mean ROC AUC = 0.9092

Plotting ROC Curves for the models

from sklearn.metrics import roc_curve, auc
import matplotlib.pyplot as plt

for name, model in models.items():
    model.fit(X_train, y_train)
    y_pred_prob = model.predict_proba(X_test)[:,1]
    fpr, tpr, thresholds = roc_curve(y_test, y_pred_prob)
    roc_auc = auc(fpr, tpr)
    plt.plot(fpr, tpr, label=f'{name} (area = {roc_auc:.2f})')

plt.plot([0, 1], [0, 1],'r--')
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.05])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('Receiver Operating Characteristic')
plt.legend(loc="lower right")
plt.show()

Confusion Matrix

from sklearn.metrics import confusion_matrix

for name, model in models.items():
    model.fit(X_train, y_train)
    y_pred = model.predict(X_test)
    print(f"{name}:\n", confusion_matrix(y_test, y_pred))

Logistic Regression:
 [[1970   15]
 [  47   14]]
Decision Tree:
 [[1936   49]
 [  52    9]]
Random Forest:
 [[1973   12]
 [  55    6]]
Gradient Boosting:
 [[1964   21]
 [  52    9]]
Support Vector Machine:
 [[1984    1]
 [  58    3]]
Neural Network:
 [[1980    5]
 [  54    7]]
XGBoost:
 [[1971   14]
 [  47   14]]
AdaBoost:
 [[1968   17]
 [  52    9]]

Classification Report

from sklearn.metrics import classification_report

for name, model in models.items():
    model.fit(X_train, y_train)
    y_pred = model.predict(X_test)
    print(f"{name}:\n", classification_report(y_test, y_pred))

Logistic Regression:
               precision    recall  f1-score   support

           0       0.98      0.99      0.98      1985
           1       0.48      0.23      0.31        61

    accuracy                           0.97      2046
   macro avg       0.73      0.61      0.65      2046
weighted avg       0.96      0.97      0.96      2046

Decision Tree:
               precision    recall  f1-score   support

           0       0.98      0.98      0.98      1985
           1       0.23      0.25      0.24        61

    accuracy                           0.95      2046
   macro avg       0.61      0.61      0.61      2046
weighted avg       0.95      0.95      0.95      2046

Random Forest:
               precision    recall  f1-score   support

           0       0.97      0.99      0.98      1985
           1       0.37      0.11      0.17        61

    accuracy                           0.97      2046
   macro avg       0.67      0.55      0.58      2046
weighted avg       0.96      0.97      0.96      2046

Gradient Boosting:
               precision    recall  f1-score   support

           0       0.97      0.99      0.98      1985
           1       0.29      0.15      0.20        61

    accuracy                           0.96      2046
   macro avg       0.63      0.57      0.59      2046
weighted avg       0.95      0.96      0.96      2046

Support Vector Machine:
               precision    recall  f1-score   support

           0       0.97      1.00      0.99      1985
           1       0.75      0.05      0.09        61

    accuracy                           0.97      2046
   macro avg       0.86      0.52      0.54      2046
weighted avg       0.96      0.97      0.96      2046

Neural Network:
               precision    recall  f1-score   support

           0       0.98      0.99      0.98      1985
           1       0.48      0.20      0.28        61

    accuracy                           0.97      2046
   macro avg       0.73      0.60      0.63      2046
weighted avg       0.96      0.97      0.96      2046

XGBoost:
               precision    recall  f1-score   support

           0       0.98      0.99      0.98      1985
           1       0.50      0.23      0.31        61

    accuracy                           0.97      2046
   macro avg       0.74      0.61      0.65      2046
weighted avg       0.96      0.97      0.96      2046

AdaBoost:
               precision    recall  f1-score   support

           0       0.97      0.99      0.98      1985
           1       0.35      0.15      0.21        61

    accuracy                           0.97      2046
   macro avg       0.66      0.57      0.59      2046
weighted avg       0.96      0.97      0.96      2046

The classification reports are particularly useful in our case as the data is very strongly skewed towards the negative class. The classification report gives us a good idea of how well the model is performing on the positive class.

Looking at the metrics for the positive class, we can see that Support Vector Machines are by far the best model for this dataset, outperforming the next best model by precision, XGBoost by $25\%$ and in recall/f1 score, Neural Networks by more than $20\%$.

The following are the statistics for the SVM model:

	Precision	Recall	F1-Score	Support
0	0.97	1.00	0.99	1985
1	0.75	0.05	0.09	61
Accuracy			0.97	2046
Macro Avg	0.86	0.52	0.54	2046
Weighted Avg	0.96	0.97	0.96	2046

Plotting Learning Curves

import numpy as np
import matplotlib.pyplot as plt
from sklearn.model_selection import learning_curve

def plot_learning_curve(ax, estimator, title, X, y, ylim=None, cv=None, n_jobs=None, train_sizes=np.linspace(.1, 1.0, 10)):
    ax.set_title(title)
    if ylim is not None:
        ax.set_ylim(*ylim)
    ax.set_xlabel("Training examples")
    ax.set_ylabel("Score")
    train_sizes, train_scores, test_scores = learning_curve(estimator, X, y, cv=cv, n_jobs=n_jobs, train_sizes=train_sizes)
    train_scores_mean = np.mean(train_scores, axis=1)
    train_scores_std = np.std(train_scores, axis=1)
    test_scores_mean = np.mean(test_scores, axis=1)
    test_scores_std = np.std(test_scores, axis=1)
    ax.grid()

    ax.fill_between(train_sizes, train_scores_mean - train_scores_std, train_scores_mean + train_scores_std, alpha=0.1, color="r")
    ax.fill_between(train_sizes, test_scores_mean - test_scores_std, test_scores_mean + test_scores_std, alpha=0.1, color="g")
    ax.plot(train_sizes, train_scores_mean, 'o-', color="r", label="Training score")
    ax.plot(train_sizes, test_scores_mean, 'o-', color="g", label="Cross-validation score")

    ax.legend(loc="best")

# Calculate the number of rows needed for subplots
n_models = len(models)
n_cols = 2
n_rows = n_models // n_cols + (n_models % n_cols > 0)

fig, axes = plt.subplots(n_rows, n_cols, figsize=(15, n_rows * 5))

# Flatten the axes array and remove extra subplots
axes = axes.flatten()[:n_models]

# Loop over your models
for ax, (name, model) in zip(axes, models.items()):
    title = f"Learning Curves ({name})"
    cv = 10
    plot_learning_curve(ax, model, title, X_train, y_train, cv=cv, n_jobs=4)

# Remove empty subplots
for ax in axes[n_models:]:
    ax.remove()

plt.tight_layout()
plt.show()

Looks like SVM's are the way to go.

## Using the SVM model on the data and hyper-parameter tuning

from sklearn.model_selection import GridSearchCV

param_grid = {
    'C': [0.1, 1, 10, 100],
    'gamma': [1, 0.1, 'scale', 'auto'],
    'kernel': ['rbf', 'poly', 'sigmoid']
}

grid = GridSearchCV(SVC(class_weight='balanced'), param_grid, refit=True, verbose=2, cv=5, scoring='roc_auc')
grid.fit(X_train, y_train)

print(grid.best_params_)
print(grid.best_estimator_)
print(grid.best_score_)

# Using the best parameters to predict the test data
grid_predictions = grid.predict(X_test)

print(confusion_matrix(y_test, grid_predictions))
print(classification_report(y_test, grid_predictions))

Fitting 5 folds for each of 48 candidates, totalling 240 fits
[CV] END .........................C=0.1, gamma=1, kernel=rbf; total time=   0.6s
[CV] END .........................C=0.1, gamma=1, kernel=rbf; total time=   0.5s
[CV] END .........................C=0.1, gamma=1, kernel=rbf; total time=   0.5s
[CV] END .........................C=0.1, gamma=1, kernel=rbf; total time=   0.5s
[CV] END .........................C=0.1, gamma=1, kernel=rbf; total time=   0.5s
[CV] END ........................C=0.1, gamma=1, kernel=poly; total time=   6.8s
[CV] END ........................C=0.1, gamma=1, kernel=poly; total time=   7.2s
[CV] END ........................C=0.1, gamma=1, kernel=poly; total time=   6.6s
[CV] END ........................C=0.1, gamma=1, kernel=poly; total time=   5.6s
[CV] END ........................C=0.1, gamma=1, kernel=poly; total time=   5.9s
[CV] END .....................C=0.1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=0.1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=0.1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=0.1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=0.1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=0.1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=0.1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=0.1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=0.1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=0.1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.1, kernel=poly; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.1, kernel=poly; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.1, kernel=poly; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.1, kernel=poly; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.1, kernel=poly; total time=   0.1s
[CV] END ...................C=0.1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=0.1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=0.1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=0.1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=0.1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=0.1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .....................C=0.1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END .....................C=0.1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END .....................C=0.1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END .....................C=0.1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END .....................C=0.1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END ..................C=0.1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=0.1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=0.1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=0.1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=0.1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=0.1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=0.1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=0.1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=0.1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=0.1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ....................C=0.1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ....................C=0.1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ....................C=0.1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ....................C=0.1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ....................C=0.1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END .................C=0.1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END .................C=0.1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END .................C=0.1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END .................C=0.1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END .................C=0.1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END ...........................C=1, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ...........................C=1, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ...........................C=1, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ...........................C=1, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ...........................C=1, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ..........................C=1, gamma=1, kernel=poly; total time=  25.5s
[CV] END ..........................C=1, gamma=1, kernel=poly; total time=  26.1s
[CV] END ..........................C=1, gamma=1, kernel=poly; total time=  19.7s
[CV] END ..........................C=1, gamma=1, kernel=poly; total time=  29.8s
[CV] END ..........................C=1, gamma=1, kernel=poly; total time=  28.6s
[CV] END .......................C=1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=1, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .........................C=1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .........................C=1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .........................C=1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .........................C=1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .........................C=1, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ........................C=1, gamma=0.1, kernel=poly; total time=   0.2s
[CV] END ........................C=1, gamma=0.1, kernel=poly; total time=   0.2s
[CV] END ........................C=1, gamma=0.1, kernel=poly; total time=   0.2s
[CV] END ........................C=1, gamma=0.1, kernel=poly; total time=   0.3s
[CV] END ........................C=1, gamma=0.1, kernel=poly; total time=   0.2s
[CV] END .....................C=1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=1, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ........................C=1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ........................C=1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ........................C=1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ........................C=1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ........................C=1, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .......................C=1, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END .......................C=1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END .......................C=1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END .......................C=1, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END .......................C=1, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END ....................C=1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=1, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .......................C=1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .......................C=1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .......................C=1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .......................C=1, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ......................C=1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ......................C=1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ......................C=1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ......................C=1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ......................C=1, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ...................C=1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END ...................C=1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END ...................C=1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END ...................C=1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END ...................C=1, gamma=0.001, kernel=sigmoid; total time=   0.0s
[CV] END ..........................C=10, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ..........................C=10, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ..........................C=10, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ..........................C=10, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ..........................C=10, gamma=1, kernel=rbf; total time=   0.6s
[CV] END .........................C=10, gamma=1, kernel=poly; total time= 1.3min
[CV] END .........................C=10, gamma=1, kernel=poly; total time=  39.9s
[CV] END .........................C=10, gamma=1, kernel=poly; total time=  53.7s
[CV] END .........................C=10, gamma=1, kernel=poly; total time= 1.1min
[CV] END .........................C=10, gamma=1, kernel=poly; total time= 1.2min
[CV] END ......................C=10, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=10, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=10, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=10, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=10, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END ........................C=10, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ........................C=10, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ........................C=10, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ........................C=10, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ........................C=10, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=10, gamma=0.1, kernel=poly; total time=   1.2s
[CV] END .......................C=10, gamma=0.1, kernel=poly; total time=   1.1s
[CV] END .......................C=10, gamma=0.1, kernel=poly; total time=   1.0s
[CV] END .......................C=10, gamma=0.1, kernel=poly; total time=   1.1s
[CV] END .......................C=10, gamma=0.1, kernel=poly; total time=   1.4s
[CV] END ....................C=10, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=10, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=10, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=10, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ....................C=10, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=10, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .......................C=10, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .......................C=10, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .......................C=10, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .......................C=10, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=10, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END ......................C=10, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END ......................C=10, gamma=0.01, kernel=poly; total time=   0.0s
[CV] END ......................C=10, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END ......................C=10, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END ...................C=10, gamma=0.01, kernel=sigmoid; total time=   0.0s
[CV] END ...................C=10, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=10, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=10, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=10, gamma=0.01, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=10, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ......................C=10, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ......................C=10, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ......................C=10, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ......................C=10, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=10, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END .....................C=10, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END .....................C=10, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END .....................C=10, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END .....................C=10, gamma=0.001, kernel=poly; total time=   0.0s
[CV] END ..................C=10, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=10, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=10, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=10, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END ..................C=10, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END .........................C=100, gamma=1, kernel=rbf; total time=   0.6s
[CV] END .........................C=100, gamma=1, kernel=rbf; total time=   0.6s
[CV] END .........................C=100, gamma=1, kernel=rbf; total time=   0.6s
[CV] END .........................C=100, gamma=1, kernel=rbf; total time=   0.6s
[CV] END .........................C=100, gamma=1, kernel=rbf; total time=   0.6s
[CV] END ........................C=100, gamma=1, kernel=poly; total time= 1.4min
[CV] END ........................C=100, gamma=1, kernel=poly; total time=  59.4s
[CV] END ........................C=100, gamma=1, kernel=poly; total time=  42.2s
[CV] END ........................C=100, gamma=1, kernel=poly; total time= 1.1min
[CV] END ........................C=100, gamma=1, kernel=poly; total time= 1.3min
[CV] END .....................C=100, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=100, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=100, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=100, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .....................C=100, gamma=1, kernel=sigmoid; total time=   0.1s
[CV] END .......................C=100, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=100, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=100, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=100, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END .......................C=100, gamma=0.1, kernel=rbf; total time=   0.1s
[CV] END ......................C=100, gamma=0.1, kernel=poly; total time=   6.9s
[CV] END ......................C=100, gamma=0.1, kernel=poly; total time=   7.2s
[CV] END ......................C=100, gamma=0.1, kernel=poly; total time=   6.9s
[CV] END ......................C=100, gamma=0.1, kernel=poly; total time=   6.2s
[CV] END ......................C=100, gamma=0.1, kernel=poly; total time=   5.5s
[CV] END ...................C=100, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=100, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=100, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=100, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ...................C=100, gamma=0.1, kernel=sigmoid; total time=   0.1s
[CV] END ......................C=100, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=100, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=100, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=100, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END ......................C=100, gamma=0.01, kernel=rbf; total time=   0.1s
[CV] END .....................C=100, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END .....................C=100, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END .....................C=100, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END .....................C=100, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END .....................C=100, gamma=0.01, kernel=poly; total time=   0.1s
[CV] END ..................C=100, gamma=0.01, kernel=sigmoid; total time=   0.0s
[CV] END ..................C=100, gamma=0.01, kernel=sigmoid; total time=   0.0s
[CV] END ..................C=100, gamma=0.01, kernel=sigmoid; total time=   0.0s
[CV] END ..................C=100, gamma=0.01, kernel=sigmoid; total time=   0.0s
[CV] END ..................C=100, gamma=0.01, kernel=sigmoid; total time=   0.0s
[CV] END .....................C=100, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=100, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=100, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=100, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END .....................C=100, gamma=0.001, kernel=rbf; total time=   0.1s
[CV] END ....................C=100, gamma=0.001, kernel=poly; total time=   0.1s
[CV] END ....................C=100, gamma=0.001, kernel=poly; total time=   0.1s
[CV] END ....................C=100, gamma=0.001, kernel=poly; total time=   0.1s
[CV] END ....................C=100, gamma=0.001, kernel=poly; total time=   0.1s
[CV] END ....................C=100, gamma=0.001, kernel=poly; total time=   0.1s
[CV] END .................C=100, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END .................C=100, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END .................C=100, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END .................C=100, gamma=0.001, kernel=sigmoid; total time=   0.1s
[CV] END .................C=100, gamma=0.001, kernel=sigmoid; total time=   0.1s
{'C': 0.1, 'gamma': 0.001, 'kernel': 'poly'}
SVC(C=0.1, gamma=0.001, kernel='poly')
0.925164008452842
[[1985    0]
 [  61    0]]
              precision    recall  f1-score   support

           0       0.97      1.00      0.98      1985
           1       0.00      0.00      0.00        61

    accuracy                           0.97      2046
   macro avg       0.49      0.50      0.49      2046
weighted avg       0.94      0.97      0.96      2046

/Users/arvindtawker/miniconda3/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))
/Users/arvindtawker/miniconda3/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))
/Users/arvindtawker/miniconda3/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))